mlsanityData to 7 Oct 2026

DeepEyesV2: Toward Agentic Multimodal Model

arXiv 2511.05271 ↗First posted 7 Nov 2025cs.CV cs.AIcited by 119

On the map: LLM agents and tool use → Multimodal and embodied agents

Abstract (arXiv)

Agentic multimodal models should not only comprehend text and images, but also actively invoke external tools, such as code execution environments and web search, and integrate these operations into reasoning. In this work, we introduce DeepEyesV2 and explore how to build an agentic multimodal model from the perspectives of data construction, training methods, and model evaluation. We observe that direct reinforcement learning alone fails to induce robust tool-use behavior. This phenomenon motivates a two-stage training pipeline: a cold-start stage to establish tool-use patterns, and reinforcement learning stage to further refine tool invocation. We curate a diverse, moderately challenging training dataset, specifically including examples where tool use is beneficial. We further introduce RealX-Bench, a comprehensive benchmark designed to evaluate real-world multimodal reasoning, which inherently requires the integration of multiple capabilities, including perception, search, and reasoning. We evaluate DeepEyesV2 on RealX-Bench and other representative benchmarks, demonstrating its effectiveness across real-world understanding, mathematical reasoning, and search-intensive tasks. Moreover, DeepEyesV2 exhibits task-adaptive tool invocation, tending to use image operations for perception tasks and numerical computations for reasoning tasks. Reinforcement learning further enables complex tool combinations and allows model to selectively invoke tools based on context. We hope our study can provide guidance for community in developing agentic multimodal models.

Key facts. Cited by 119 papers in the map (data through 7 Oct 2026). Placed in Multimodal and embodied agents, one of 16 directions in LLM agents and tool use. Builds on 56 of its 65 references that are in the map. First posted on arXiv 7 Nov 2025.

Builds on

56 of 65 references.

Closest ancestors · 4

Same direction, newest first

  1. 2025TIR-Bench: A Comprehensive Benchmark for Agentic Thinking-with-Images Reasoning3 Nov
  2. 2025WebWatcher: Breaking New Frontier of Vision-Language Deep Research Agent7 Aug
  3. 2025PyVision: Agentic Vision with Dynamic Tooling10 Jul
  4. 2025Visual Agentic Reinforcement Fine-Tuning20 May

Other references · 52

  1. 2023Visual Instruction Tuningcited by 7,792
  2. 2023BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Modelscited by 6,069
  3. 2024GPT-4o System Cardcited by 4,467
  4. 2023Improved Baselines with Visual Instruction Tuningcited by 4,362
  5. 2023Qwen Technical Reportcited by 3,273
  6. 2025DAPO: An Open-Source LLM Reinforcement Learning System at Scalecited by 2,678
  7. 2024LLaVA-OneVision: Easy Visual Task Transfercited by 2,669
  8. 2023InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Taskscited by 2,534
  9. 2025InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Modelscited by 1,769
  10. 2023MathVista: Evaluating Mathematical Reasoning of Foundation Models in Visual Contextscited by 1,582
  11. 2022ChartQA: A Benchmark for Question Answering about Charts with Visual and Logical Reasoningcited by 1,449
  12. 2025Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learningcited by 1,371
38 more
  1. 2023SEED-Bench: Benchmarking Multimodal LLMs with Generative Comprehensioncited by 929
  2. 2024MathVerse: Does Your Multi-modal LLM Truly See the Diagrams in Visual Math Problems?cited by 573
  3. 2024Measuring Multimodal Mathematical Reasoning with MATH-Vision Datasetcited by 516
  4. 2023OCRBench: On the Hidden Mystery of OCR in Large Multimodal Modelscited by 466
  5. 2023V*: Guided Visual Search as a Core Mechanism in Multimodal LLMscited by 459
  6. 2024VLMEvalKit: An Open-Source Toolkit for Evaluating Large Multi-Modality Modelscited by 443
  7. 2025DeepEyes: Incentivizing "Thinking with Images" via Reinforcement Learningcited by 398
  8. 2025ReTool: Reinforcement Learning for Strategic Tool Use in LLMscited by 377
  9. 2025R1-Searcher: Incentivizing the Search Capability in LLMs via Reinforcement Learningcited by 322
  10. 2025VL-Rethinker: Incentivizing Self-Reflection of Vision-Language Models with Reinforcement Learningcited by 315
  11. 2024MME-RealWorld: Could Your Multimodal LLM Challenge High-Resolution Real-World Scenarios that are Difficult for Humans?cited by 284
  12. 2024We-Math: Does Your Large Multimodal Model Achieve Human-like Mathematical Reasoning?cited by 281
  13. 2025DeepResearcher: Scaling Deep Research via Reinforcement Learning in Real-world Environmentscited by 268
  14. 2025Pixel Reasoner: Incentivizing Pixel-Space Reasoning with Curiosity-Driven Reinforcement Learningcited by 267
  15. 2024LogicVista: Multimodal LLM Logical Reasoning Benchmark in Visual Contextscited by 240
  16. 2024CharXiv: Charting Gaps in Realistic Chart Understanding in Multimodal LLMscited by 229
  17. 2025SFT or RL? An Early Investigation into Training R1-Like Reasoning Large Vision-Language Modelscited by 209
  18. 2023Can Pre-trained Vision and Language Models Answer Visual Information-Seeking Questions?cited by 203
  19. 2024VITA: Towards Open-Source Interactive Omni Multimodal LLMcited by 172
  20. 2025Thyme: Think Beyond Imagescited by 160
  21. 2023MathCoder: Seamless Code Integration in LLMs for Enhanced Mathematical Reasoningcited by 155
  22. 2018TallyQA: Answering Complex Counting Questionscited by 152
  23. 2024Divide, Conquer and Combine: A Training-Free Framework for High-Resolution Image Perception in Multimodal Large Language Modelscited by 151
  24. 2025WorldSense: Evaluating Real-world Omnimodal Understanding for Multimodal LLMscited by 147
  25. 2024SEED-Bench-2-Plus: Benchmarking Multimodal Large Language Models with Text-Rich Visual Comprehensioncited by 144
  26. 2025SoTA with Less: MCTS-Guided Sample Selection for Data-Efficient Visual Reasoning Self-Improvementcited by 144
  27. 2024Multimodal ArXiv: A Dataset for Improving Scientific Comprehension of Large Vision-Language Modelscited by 138
  28. 2025GRIT: Teaching MLLMs to Think with Imagescited by 121
  29. 2025MMSearch-R1: Incentivizing LMMs to Searchcited by 120
  30. 2025Mini-o3: Scaling Up Reasoning Patterns and Interaction Turns for Visual Searchcited by 116
  31. 2025Baichuan-Omni-1.5 Technical Reportcited by 107
  32. 2025R1-Omni: Explainable Omni-Multimodal Emotion Recognition with Reinforcement Learningcited by 99
  33. 2024MMSearch: Benchmarking the Potential of Large Models as Multi-modal Search Enginescited by 98
  34. 2025Adaptive Chain-of-Focus Reasoning via Dynamic Visual Search and Zooming for Efficient VLMscited by 94
  35. 2025VGR: Visual Grounded Reasoningcited by 70
  36. 2025VRAG-RL: Empower Vision-Perception-Based RAG for Visually Rich Information Understanding via Iterative Reasoning with Reinforcement Learningcited by 68
  37. 2025Traceable Evidence Enhanced Visual Grounded Reasoning: Evaluation and Methodologycited by 67
  38. 2025SimpleVQA: Multimodal Factuality Evaluation for Multimodal Large Language Modelscited by 61

50 most cited of 52 All 52 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 9; 2026: 110.

119 papers.

Sort:

By area:

  1. 2026XSkill: Continual Learning from Experience and Skills in Multimodal Agentscited by 71
  2. 2026Vision-OPD: Learning to See Fine Details for Multimodal LLMs via On-Policy Self-Distillationcited by 71
  3. 2026Zooming without Zooming: Region-to-Image Distillation for Fine-Grained Multimodal Perceptioncited by 47
  4. 2026REDSearcher: A Scalable and Cost-Efficient Framework for Long-Horizon Search Agentscited by 43
  5. 2025SenseNova-MARS: Empowering Multimodal Agentic Reasoning and Search via Reinforcement Learningcited by 31
  6. 2026Vision-DeepResearch Benchmark: Rethinking Visual and Textual Search for Multimodal Large Language Modelscited by 30
  7. 2025Reasoning Within the Mind: Dynamic Multimodal Interleaving in Latent Spacecited by 25
  8. 2026AgentVista: Evaluating Multimodal Agents in Ultra-Challenging Realistic Visual Scenarioscited by 22
  9. 2026GenAgent: Scaling Text-to-Image Generation via Agentic Multimodal Reasoningcited by 20
  10. 2026Video-o3: Native Interleaved Clue Seeking for Long Video Multi-Hop Reasoningcited by 20
  11. 2025DiffThinker: Towards Generative Multimodal Reasoning with Diffusion Modelscited by 18
  12. 2026MM-DeepResearch: A Simple and Effective Multimodal Agentic Search Baselinecited by 18
38 more
  1. 2026OpenSearch-VL: An Open Recipe for Frontier Multimodal Search Agentscited by 18
  2. 2026PyVision-RL: Forging Open Agentic Vision Models via RLcited by 14
  3. 2026Act Wisely: Cultivating Meta-Cognitive Tool Use in Agentic Multimodal Modelscited by 13
  4. 2026VTC-Bench: Evaluating Agentic Multimodal Models via Compositional Visual Tool Chainingcited by 12
  5. 2026Imagination Helps Visual Reasoning, But Not Yet in Latent Spacecited by 10
  6. 2026ToolCUA: Towards Optimal GUI-Tool Path Orchestration for Computer Use Agentscited by 10
  7. 2026V-Zero: Answer-Label-Free On-Policy Distillation with Contrastive Evidence Gating for Fine-Grained Visual Reasoningcited by 10
  8. 2026What Does Vision Tool-Use Reinforcement Learning Really Learn? Disentangling Tool-Induced and Intrinsic Effects for Crop-and-Zoomcited by 9
  9. 2026GeoEyes: On-Demand Visual Focusing for Evidence-Grounded Understanding of Ultra-High-Resolution Remote Sensing Imagerycited by 9
  10. 2026Towards Long-horizon Agentic Multimodal Searchcited by 9
  11. 2025Deep But Reliable: Advancing Multi-turn Reasoning for Thinking with Imagescited by 8
  12. 2026VisBrowse-Bench: Benchmarking Visual-Native Search for Multimodal Browsing Agentscited by 8
  13. 2026ResAdapt: Adaptive Resolution for Efficient Multimodal Reasoningcited by 8
  14. 2026HyperEyes: Dual-Grained Efficiency-Aware Reinforcement Learning for Parallel Multimodal Search Agentscited by 8
  15. 2026OPD-V: Visual On-Policy Self-Distillation with Modality Balancecited by 7
  16. 2026Seg-ReSearch: Segmentation with Interleaved Reasoning and External Searchcited by 6
  17. 2026VideoTemp-o3: Harmonizing Temporal Grounding and Video Understanding in Agentic Thinking-with-Videoscited by 6
  18. 2026MedXIAOHE: A Comprehensive Recipe for Building Medical MLLMscited by 6
  19. 2026VSearcher: Long-Horizon Multimodal Search Agent via Reinforcement Learningcited by 6
  20. 2026Towards On-Policy Data Evolution for Visual-Native Multimodal Deep Search Agentscited by 6
  21. 2025RSAgent: Learning to Reason and Act for Text-Guided Segmentation via Multi-Turn Tool Invocationscited by 5
  22. 2026BrowseComp-V3: A Visual, Vertical, and Verifiable Benchmark for Multimodal Browsing Agentscited by 5
  23. 2026AgenticOCR: Parsing Only What You Need for Efficient Retrieval-Augmented Generationcited by 5
  24. 2026Meissa: Multi-modal Medical Agentic Intelligencecited by 5
  25. 2026Improving Visual Reasoning with Iterative Evidence Refinementcited by 5
  26. 2026Q-Zoom: Query-Aware Adaptive Perception for Efficient Multimodal Large Language Modelscited by 5
  27. 2026Visual-Seeker: Towards Visual-Native Multimodal Agentic Search via Active Visual Reasoningcited by 5
  28. 2026VideoSearcher: Empowering Video Deep Research with Multi-Tool Agentic Reasoning via Reinforcement Learningcited by 5
  29. 2026VAD: Attributing Visual Evidence for Target Reconstruction in Multimodal On-Policy Distillationcited by 5
  30. 2025SciEGQA: A Dataset for Scientific Evidence-Grounded Question Answering and Reasoningcited by 4
  31. 2025Training Multi-Image Vision Agents via End2End Reinforcement Learningcited by 4
  32. 2025Code-in-the-Loop Forensics: Agentic Tool Use for Image Forgery Detectioncited by 4
  33. 2026SwimBird: Eliciting Switchable Reasoning Mode in Hybrid Autoregressive MLLMscited by 4
  34. 2026TikArt: Stabilizing Aperture-Guided Fine-Grained Visual Reasoning with Reinforcement Learningcited by 4
  35. 2026StreamingClaw Technical Reportcited by 4
  36. 2026Agentic-MME: What Agentic Capability Really Brings to Multimodal Intelligence?cited by 4
  37. 2026Walk the Talk: Bridging the Reasoning-Action Gap for Thinking with Images via Multimodal Agentic Policy Optimizationcited by 4
  38. 2026Thinking with Imagination: Agentic Visual Spatial Reasoning with World Simulatorscited by 4

50 most cited of 119 All 119 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2025Agent-X: Evaluating Deep Multimodal Reasoning in Vision-Centric Agentic Tasks
  2. 2025Skywork-R1V4: Toward Agentic Multimodal Intelligence through Interleaved Thinking with Images and DeepResearch
  3. 2026ToolArtist: Tool-Using Unified Multimodal Models for Agentic Image Generation
  4. 2026MetaForge: A Self-Evolving Multimodal Agent that Retrieves, Adapts, and Forges Tools On Demand ↗
  5. 2026Do Multimodal Agents Really Benefit from Tool Use? A Systematic Study of Capability Gains
  6. 2026A Survey on Foundations and Frontiers of Multimodal Agentic Frameworks: Techniques and Applications
  7. 2026VC-Tooler: Learning Compositional and Adaptive Visual Tool Use
  8. 2026GLM-5V-Turbo: Toward a Native Foundation Model for Multimodal Agents